In complex program systems according to the prior art, functions for accessing other components are generally implemented by way of extensive special program codes within the individual program objects. This is the case, for example, when specific program objects access background jobs which are expected to service specific tasks for these program objects. The code for addressing and administrating the job program objects, which then run in the background, is implemented in the calling program objects, while each individual instance of the job program objects has a dedicated queue for buffering information which must be processed.
Other functionalities, such as e.g. the monitoring of a desired maximal number of concurrently active instances of a program object (e.g. in order to ensure that the computer on which the programs run is not overloaded by too many instances of the program object and consequently impaired in terms of its responsiveness), also take place in the program object itself by way of suitable counting mechanisms, etc. As a result of all this, the individual program objects of a system are overloaded with administration codes which, being specifically adapted to the individual program objects in most cases, are moreover not replicable and therefore also increase the number of possible error sources within the overall system as a result of continuous reprogramming.